How Trump Flipped Pennsylvania
There’s lots of collective WTF on the left about the election. Trump won, but how did that happen? There are lots of theories. Those theories include, in no particular order or level of plausibility:
1. Republican state governments suppressing Democratic votes in swing states
2. Apathetic Democrats staying home
3. Trump surge (a truly silent plurality whom the polls didn’t pick up but who came out to vote)
A look at the results for Pennsylvania, especially county by county, is instructive. Pennsylvania has voted Democratic in 7 of the last 10 presidential elections, including every election from 1992 through 2012. Pennsylvania hasn’t been Republican as an electoral state since the Reagan years. So what happened?
Trump Flipped Pennsylvania
We know this but it’s worth stating anyway. In 2012, Obama won PA by about 284,000 votes. In 2016, Trump won PA by about 68,000 votes, a swing of 352,000 votes.
The Map Didn’t Change Much
At a county level, most Pennsylvania counties voted the same way in 2016 as in 2012. Clinton “flipped” two of Romney’s counties, and Trump “flipped” three of Obama’s. It doesn’t make much sense to talk about flipping counties, since winning a county, unlike winning a state, is meaningless in itself. But some of the flips point to interesting changes, as we’ll see.
More Voters Came Out
Overall, turnout in PA rose from 5,445,000 in 2012 to 5,757,000 in 2016, an increase of 312,000, or about 6%.
Most of the New Voters Voted For Trump
Relative to 2012, Trump captured 332,000 more votes than Romney, while Clinton captured 20,000 FEWER votes than Obama. One way to read that, and the way that seems most likely, is that the vast majority of those 312,000 new votes went to Trump. The county-by-county numbers add some nuance, but as a broad-brush view, this seems not too inaccurate.
The Democrats Increased Their Base And Flipped Some More-Affluent Republicans …
Clinton actually drew significantly more votes than Obama in the counties that she won. She won 11 of the state’s 67 counties (whereas Obama won 12, though Clinton flipped 2 counties and Trump flipped 3). In the Democratic-won counties, which cluster around urban areas like Pittsburgh, Philadelphia, Harrisburg, State College, Allentown, Scranton and Wilkes-Barre, she improved on Obama’s vote totals by 5.3%. Her most significant achievement, perhaps, was to flip Chester county, which is outside Philadelphia in the “next ring” of counties (I actually live there). Romney narrowly won Chester in 2012 by about 1900 votes. Clinton surged 22% there in 2016 and won it by 25,000 votes. This suggests strongly that relatively affluent and probably moderate Republicans in Chester were appalled by Trump and flipped.
But Trump Vastly Increased His Base And Flipped Some Less-Affluent Democrats
As noted, Trump improved on Romney’s vote total by 332,000 votes, or about 13% on average. But in the counties he won, he improved on Romney’s numbers by 18.5%. This was a broad-based surge across all the 56 counties he won. Clinton had one Democratic county, Chester, where she improved on 2012 by 20% or more. Trump did the same thing in 19 counties. Significantly, just as Clinton flipped Chester county, Trump flipped Luzerne county, just northwest of Wilkes-Barre, improving on Romney by 19,000 votes, more or less canceling Clinton’s additional 25,000 in Chester. Luzerne is certainly more middle-class and working-class than Chester, and obviously a large number of these voters rejected the Democrat this time around.
The won-lost map in PA didn’t change much at the county level. The biggest events, the flips of Chester and Luzerne counties, show clearly the new lines of polarization, though these events mostly canceled each other. So the map only changed slightly, and both candidates increased their party’s vote totals in the counties they won. But whereas Clinton boosted the 2012 numbers by 6%, Trump improved on the 2012 numbers by 13%. Clinton lost voters in all but four counties that she didn’t win outright. Trump lost voters in only one county, PERIOD. That means that even in counties Clinton won, Trump improved on Romney’s numbers. In Lackawanna county, for example, Trump surged by 40%, almost flipping the entire county; he surged 26% in Monroe county, which Clinton won by a mere 224 votes. Trump added 16% to the 2012 totals in Philadelphia county.
The bottom line is that Trump added voters in all but one county. All but one county!
Clinton didn’t need much. She needed another 65,000 votes. But she lost ground relative to Obama in 54 of 67 counties,while Trump gained a lot of ground relative to Romney in 66 out 67.
That is a pretty damn broad-based revolt.
If you’re interested, here’s an Excel file with a county comparison of the 2012 and 2016 elections in PA.